Leukemia Subtype Classification via Gene Expression Using Machine Learning Algorithms

This project aims to improve patient care and treatment outcomes by using gene expression data to reliably diagnose acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL). To create trustworthy classification models, six machine learning (ML) techniques were used: Gaussian Naive Bayes (NB), Random Forest, Support Vector Machine (SVM), K Nearest Neighbors (KNN), Multi-Layer Perceptron(MLP), and Logistic Regression. The experiment's outcomes show that the algorithms were successful; 91% accuracy rates were achieved by both logistic regression and MLP, while the other approaches produced results that were competitive. Good findings for the ROC curve analysis area under the curve (ROCAUC) demonstrated the models' ability to distinguish between different forms of cancer. These results demonstrate the application of ML algorithms in oncology.